Mapping of Data Schemas to operations and vice versa

[egekorkan]

The comment at #1324 (comment) made me realize that we do not explain very well how data schemas of an affordance maps to the operations. We implicitly know this, i.e. subscription in events maps to the subscribeevent operation. However, I think that the spec needs a nice table that says that for this operation, the data schema (or the payload) is defined by:

• these terms in the affordance
• protocol related information in the forms
• lack of both means empty payload

[benfrancis]

As part of of this, as noted in #1324 there is currently no explanation of how to define schemas for some of the meta-interactions in top level forms, e.g. subscribeallevents, unsubscribeallevents and queryallactions.

I think we need to go through every operation type and ensure it's clear how to define a schema for that operation. See also: #1053.

[benfrancis]

Something else that I noticed today while reviewing w3c/wot-profile#192 is that there's currently no way to define a schema for the response to a subscribeevent operation. Specifically in this case the need is to provide a subscription ID in the response to the creation of a webhook subscription, which the Consumer can later use to cancel the subscription.

Related to #1053.

[egekorkan]

The hacky way to do that is to use oneOf in conjunction with readOnly and writeOnly or simply indicating the read/write options per property:https://w3c.github.io/wot-thing-description/#webhook-example-serialization